For example, imagine my TIMECUBE PAC ponies up for 2500 targeted facebook micro-campaigns with very slight variations depending on the target demo with an ad spend of $20, $50K cost. Would it be actually possible for anybody to get Facebook to tell them who I was advertising at, or what those ads were?
However, this search box doesn't adequately explain who is seeing these ads. It does expose which state the ads are shown within, and the general demo (age, gender), but if I am not mistaken the ad targeting service is way more fine grained than this. Isn't this stuff down to the zip+4 or even, "near a Starbucks"? Facebook faced a lawsuit[0] in 2019 for Fair Housing Act violations, the complaint alleges that facebook let people advertise housing while excluding "Hispanic", "New Parents", etc in the target demographics.
It also looks like I can search back 7, 60, 30, 90 days or "All Impressions". There's no way a librarian had any input in this.
In addition, I suspect that if I were to actually do analysis here Facebook would start demanding PII, especially if I start crawling the searches. Is there an API for this? The infinite scrolling is IMO useless for any real use cases such as monitoring political ads in general.
It does appears as if I could use this search box to find one ad campaign or all of one org's campaigns. But I kind of suspect I'd need a whole bunch of grad students to do data collection?
Do you know if other platforms also expose some political ads?
[0] https://www.salon.com/2019/08/21/digital-redlining-facebooks...
TV ads are broadcast and you can have one grad student check some boxes per market, they just need an extension cord and a bunch of TVs to watch.
Internet ads can have a spend of one and they're gone, or behind this really slick infinite scroll. If 40 people see the ad really whats the likelihood of gleaning useful information?
The Old World Mad Men style Ad is blasted from the antenna, printed, or plastered on a billboard. With Old Media you can tell easily who's being targeted because you can't narrow down based on "Likely to be Pregnant". A TV commercial is incapable of pranking your roommate[0], this stuff is a wholly new category, and its profitability is larger than Uber's 2019 Q2 posted losses.
While this handy search page is for sure an improvement over Facebook's previously completely opaque ad market, I disagree that this is an improvement over the Old Media.
I'd also imagine the Old Media ads have a lawyer or compliance person overseeing the business, and while I know FB has the lawyers I doubt anyone is motivated or empowered to hit the brakes.
[0] https://ghostinfluence.com/the-ultimate-retaliation-pranking...
Just look at the content surrounding the ads on daytime soaps, college ball, anime cartoons. Billboard next to (My apologies for this shorthand, I can't think of any polite way to make this next analogy) "Inner City" liquor store, suburban daycare, or just along the interstate.
As well as dead reckoning, I'll bet a local library has the spend on paper in some industry publication. Anyone in the industry can probably rattle off the spend and get pretty close. Plus, since they are marketing ad space for money, why not send over a proposal and see what the initial offer is?
Surely this takes more time than glancing at some Facebook reactjs ad timeline which is for sure not edited by an algorithm, but like I said there is a finite amount of billboards and tv channels. One grad student wouldn't do it, but however many you need for the internet is sure going to cost you.
a. Congressional Republican majority b. Some elected Republicans c. A group of Republican voters
If AOC ran that ad, it very well would be truthful for B and C, but should Facebook interpret the ad as describing A? There's way too much nuance to be the single arbitrator.